Results of Shoulder Arthroplasty in Primary Gleno-humeral Osteoarthritis

Summary

This prospective multicentrer study involved 155 shoulder arthroplasties operated on for osteoarthritis. Mean age at surgery was 66.6 years. The average follow-up was 19.4 months (range 12–46 months). There was a full-thickness degenerative cuff tear in 12.5 and an attempt to repair was made in only five cases. The average normalised Constant Score was 39.4% preoperatively and 95.5% at review. The results were excellent or good in 80.5%, fair in 12.8% and poor in only 6.7%. Two of the 13 pressfit humeral components demonstrated progressive radiographic loosening, whereas no signs of loosening were found in the 142 cemented humeral components. Glenoid lucencies were present in 60% of the cases, they were incomplete in 55% and progressive in 15%. Three glenoid were definitively loosed with signs of migration. We did not find any statistical correlation between the final results and a preoperative cuff tear, associated acromioplasty or associated lengthening of the subscapularis. Fourteen patients did not have a glenoid component inserted because the glenoid surface was judged to be only slightly eroded. The absence of a glenoid component significantly worsened the results: the Constant score fells from 97% to 81% in the absence of a glenoid. The poorest result observed with hemiarthroplasty was clearly related to progressive glenoid erosion. Total shoulder arthroplasty gives better functional result than hemiarthroplasty in osteoarthritis. However, loosening of the glenoid component remains a potential complication and long term radiographic control is necessary.